[Untitled]

Purpose Angular stable implants reduced the complication rate in the treatment of humeral head fractures. But the failure rate is still high. To further reduce the risk of cut-out, cement augmentation of screws was introduced. A reason for failure of plate osteosynthesis might be the extremely high stiffness of the screw-plate interface leading to a loss of reduction and cut-out of screws. A more homogeneous distribution of the forces on all screws may avoid secondary dislocation. We hypothesize that dynamic osteosynthesis minimizes screw loosening and results in a higher load to failure than standard locking screws. Methods Twelve paired human humerus specimens were analysed. A standardized three-part fracture model with a metaphyseal defect was simulated. Within each pair of humeri, one was fixed with a Philos plate and standard locking screws (LS), whereas the other humerus was fixed with a Philos plate and dynamic locking screws (DLS). A cyclic varus-bending test or a rotation test with increasing loading force was performed until failure of the screw-bone-fixation. Results In the varus bending test, pairs failed by screw loosening in the humeral head. The LS-group reached 2901 (601–5201) load cycles until failure, while the DLS-group failed after 3731 (2001–5601) cycles. This corresponds to a median loading of 195 N for the LS-group and 235 N for the DLS-group (p = 0.028). In the rotation test the LS-group reached a median of 1101 (501–1501) load cycles until failure of fixation occurred, while the DLS-group failed after 1401 (401–2201) cycles (p = 0.225). Conclusions Plate fixation using dynamic locking screws for the treatment of proximal humerus fractures demonstrated more load cycles until failure compared to standard locking plate osteosynthesis.